Journal of Medkinal Chemistry

©
Copyright 1978 by the American Chemical Society

Volume 22, Number 1 January 1979

The Benzodiazepine Story1

Leo H. Sternbach
Research Division, Hoffmann-La Roche Inc., Nutley, New Jersey 07110. Received September 5, 1978
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Before starting with my lecture proper, I would like to then known tranquilizers were intensively studied by
Downloaded via UNIV DE GUADALAJARA on November 11, 2023 at 21:33:52 (UTC).

express my deep felt thanks to the Medicinal Chemistry
Division of the American Chemical Society for choosing
me to receive this prestigious award. It is indeed an honor,
gives me great satisfaction, and makes me very happy.
In this lecture, I would like to tell you about the chain
of events that started with the synthesis of a new chemical
entity and culminated in the discovery of a new class of
biologically active agents. Specifically, I shall discuss the
development of the group of centrally acting 1,4-benzo-
diazepines that began with the discovery of a pharma-
cologically active compound, which received the generic
name chlordiazepoxide and is the active ingredient of
Librium.
The story starts in the mid 1950s when the tranquilizers,
a new class of therapeutic agents, were shown to have
considerable clinical value, and Roche decided to embark
on a program concerned with the synthesis of products of
this type. The pharmacological tests for the screening of
sedatives and tranquilizers were well in hand, and we
chemists were asked to produce a new compound which
would be superior to the then existing tranquilizers.
A chemist faced with a problem of this kind has various
approaches at his disposal. He can start in a rather so-
phisticated manner with a biochemical working hypothesis,
with other intelligent speculations, or select a more prosaic
approach.
Our knowledge of the processes occurring in the brain
was rather limited, and we could not think of an intelligent
working hypothesis. Therefore, we decided to take the low
road and to attack this problem in a purely empirical
manner. Since our main interest was chemical synthesis,
we planned to select an approach which would be
chemically most attractive, challenging, and satisfying.
This left us essentially with two alternatives: to modify
existing drugs or to search for a new class of tranquilizers.
Molecular modification, sometimes disparagingly called
molecular manipulation, of products known to have the
desired properties has proven to be very successful in the
past. The modification or simplification of the molecular
structure of naturally occurring alkaloids, hormones, and
antibiotics has given excellent results. The molecular
modification of many synthetic drugs, e.g., of the first sulfa
drug Prontosil, the first MAO inhibitor, the first synthetic
diuretic, and many other synthetic biologically active
products has also led to vastly improved medicines. This
approach did not appear to be very promising, since the
several groups of investigators, e.g., meprobamate (Mil-
town) at Wallace Laboratories, reserpine by the Ciba
research group, and chlorpromazine by the SKF research
team.
We therefore considered it more attractive to pursue the
second approach and be guided mainly by our interest in
synthetic chemical bench work. The class of compounds
we were seeking would be expected to fulfill the following
criteria: (1) be relatively unexplored, (2) be readily ac-
cessible, (3) give the possibility of a multitude of variations
and transformations, (4) offer some challenging chemical
problems, and (5) “look” as if it could lead to biologically
active products. Our search led us to the benzheptox-
diazines, compounds I had worked with during my
postdoctoral assistantship years at the University of
Cracow.2 These studies in the early 1930s were concerned
with a search for new dyestuffs and dyestuff intermediates
and were terminated after we found that the benz-
heptoxdiazines did not lend themselves to the anticipated
uses. Compounds of this type now looked rather attractive
to us and seemed to be well suited for a fairly broad
synthetic program. The starting materials were readily
accessible, and their transformation into benzheptox-
diazines seemed to be a reaction of general applicability.
The first compounds of this type were prepared in 1891
by Auwers and von Meyenburg3 by treatment of amino-
(1) or acetaminoacetophenone oximes (2) with a Beckmann
mixture. The heptoxdiazine structure was “definitely
established” in 1924.4

0022-2623/79/1822-0001$01.00/0 © 1978 American Chemical Society

2 Journal of Medicinal Chemistry, 1979, Vol. 22, No.
I knew from my past experience that these compounds
were readily formed, crystallized very well, and could be
easily isolated and purified. A literature search revealed
that since our work in Cracow2 very little had been
published about the chemistry of benzheptoxdiazines and
that no studies concerned with their biological properties
had been carried out. These compounds therefore seemed
to be ideally suited for our purposes. We planned to
synthesize a number of the relatively readily accessible
amino ketones 4 bearing various substituents in the
r2
6
benzene ring and acylate their oximes to products of type
5. By combining a variety of amino ketones and acids, a
large number of new compounds of type 6 would be ex-
pected to become available in the shortest time and with
aminimum of difficulties.
Further transformations of 6 offered the promise of
additional interesting possibilities. One of our first ob-
was the synthesis of new compounds, e.g., 7, which,
jectives
by treatment with amines, could be converted into
products possessing basic side chains as in 8. The reaction
^R
products, we hoped, might have interesting properties,
since it is known that basic groups frequently impart
biological activity.
In the midst of our work, we began to have serious
doubts about the structure of the heptoxdiazines of type
7 and 8. In particular, the results of hydrogenation ex-
periments were quite revealing. The oxygen was removed
with great ease and the products, formed in good yield,
were quinazolines. Additional chemical studies showed
unequivocally that the so-called heptoxdiazines did not
possess the postulated structure but were in fact quin-
azoline 3-oxides5 as shown in 9 and 10. The interesting
/R
R a measure
novel structure of these compounds and their facile for-
mation and transformations gave us additional incentive
to continue our work. We synthesized a number of
quinazoline 3-oxides of type 9, treated them with secondary
amines, and obtained the expected substitution products
of type 10. The reaction occurred readily with the for-
mation of nicely crystallized products, but unfortunately
1 Sternbach
the pharmacological properties were rather disappointing.
Neither removal of the Ñ-oxide oxygen nor hydrogenation
at the 3,4 position yielded anything of interest.5,6
At that time (this was the second half of 1955) we had
to stop our work in the quinazoline field since other
problems seemed to be of greater importance. We became
involved with other synthetic projects and the isolation,
purification, and degradation of various antibiotics. This
intensive work, of little practical value, finally led, in April
1957, to an almost hopeless situation. The laboratory
benches were covered with dishes, flasks, and beakers—all
containing various samples and mother liquors. The
working area had shrunk almost to zero, and a major spring
cleaning was in order.
During this cleanup operation, my co-worker, Earl
Reeder, drew my attention to a few hundred milligrams
of two products, a nicely crystalline base and its hydro-
chloride. Both the base, which had been prepared by
treating the quinazoline N-oxide 11 with methylamine, and
its hydrochloride had been made sometime in 1955. The
products were not submitted for pharmacological testing
at that time because of our involvement with other
problems. Since the compounds were pure and had the
expected composition, we submitted the water-soluble salt
for pharmacological evaluation in 1957. We again expected
to receive negative pharmacological results and thought
that our work with quinazoline N-oxides would be finished
and lead to the publication of some chemically interesting
material. Little did we know that this was the start of a
program which would keep us busy for many years.
The product was submitted for testing in May 1957 and
within a few days we received an enthusiastic telephone
call from our pharmacologist, Dr. Lowell Randall. He
informed us that this compound possessed unusually
interesting properties in the six tests which were generally
used for the preliminary screening of tranquilizers and
sedatives. Table I shows the comparison of its pharma-
cological properties with those of the then most used
tranquilizers and the hypnotic, phenobarbital.7 Mice were
used in all these tests with the exception of the third one
which was carried out with unanesthesized cats. The
inclined screen test indicates muscle relaxation and se-
dation and the foot shock possibly a taming effect. The
test with the unanesthesized cat shows muscle relaxation
and is quite characteristic for this class of compounds as
is the pentylenetetrazole test which indicates sedative and
anticonvulsant properties. The two electroshock tests are
of their potency as anticonvulsants.8
Table I shows that the new compound was much more
effective than meprobamate in each of our six preliminary
tests. Compared with chlorpromazine, it was weaker in
the first two tests, of equal strength as a muscle relaxant
in the cat, and had a more pronounced anticonvulsant
activity in the mouse. As can be seen in the last line our
new compound was superior to phenobarbital in the first
four tests but inferior in the two electroshock tests. The
absence of direct hypnotic properties below the toxic dose
was another interesting feature which differentiated it very
characteristically from phenobarbital. It is also worth
noting that unlike chlorpromazine and reserpine it had no
Award Address Journal of Medicinal Chemistry, 1979, Vol. 22, No. 1 3

Table I. Pharmacological Properties0 of “New Compound”, Meprobamate, Chlorpromazine, and Phenobarbital

anticonvulsant tests
electroshock
inclined foot pentylene-
compd screen shock cat tetrazole max min
new compound 100 40 2 18 92 150
meprobamate 250 250 100 150 200 167
chlorpromazine 17 20 2.5 42 150 600
phenobarbital 120 80 10 75 18 90
0
Dose (mg/kg) of orally administered drug required to achieve the desired effect.

Scheme I 13 via 14 and not from the quinazoline 12 or other con-

ceivable isomers.
This unusual transformation of a quinazoline 3-oxide
into a benzodiazepine 4-oxide prompted us to investigate
the reaction in depth. The study of various reaction
conditions, solvents, and the behavior of a number of
analogues6 led us to conclude that the methylamine attacks
the quinazoline N- oxide at the 2 position due to its residual
positive charge (11a) rather than replacing the reactive

nh2 HOOC
:ch2 + h2nch3
h2n
Cl •CO

15

effect whatsoever on the autonomic nervous system. The

product had a pronounced taming effect on monkeys. The
low toxicity (620 mg/kg in mice) was particularly en- chlorine atom as might be expected.10 This causes the
couraging. It looked like an ideal compound. formation of an intermediate of type 16 which most
While the compound underwent a whole gamut of so-
probably is ultimately transformed via 17 into the 1,4-
phisticated pharmacological tests by Dr. Randall and his benzodiazepine derivative 13. The interesting and
staff,7 we studied the chemistry of this unusual product. promising pharmacological properties of this compound
From the very beginning we had reservations about its led us to synthesize a number of related products (19)
structure since the UV and IR spectra were completely
different from those of the starting material 11 or those
of other related quinazoline 3-oxides. Since at that time
NMR or mass spectra were not yet used for the structure
determination of complicated heterocyclic molecules, we
resorted to classical methods. The analytical data showed
that the reaction product had the expected molecular
weight and elementary composition; the methylamino
group and the IV-oxide function were also present. The
degradative studies,9 which allowed us to establish defi-
nitely that the compound had structure 13 rather than 12, obtained by causing quinazoline 3-oxides 18 to react with
are outlined in Scheme I. ammonia and a variety of primary amines.11
First we removed the IV-oxide oxygen and then we The synthesis of these analogues and homologues en-
hydrolyzed compound 14 with acid. This gave the ami- abled us to file a patent application in May 1958 claiming
nobenzophenone used as starting material in quantitative 2-amino-1,4-benzodiazepine 4-oxides bearing various
yield. The acid-soluble residue gave, after benzoylation substituents in the benzo and phenyl rings. Because of
by the Schotten-Baumann procedure, benzoylmethyl- the novelty of these products, the patent12 was granted
amine and hippuric acid in greater than 60% yield. The within a year (July 1959) and without difficulty.
degradation products, glycine and methylamine, could The pharmacological evaluation of all the analogues and
result only from the hydrolysis of the benzodiazepinone homologues on hand showed that none were significantly
4 Journal of Medicinal Chemistry, 1979, Vol. 22, No. 1 Sternbach

Table II. Pharmacological Activity0 of Chlordiazepoxide

and Its Transformation Products
anticonvulsant tests
pen-
tylene electroshock
impel screen shock cat zole max min
13 100 40 2 18 92 150
21 100 20 2 15 150 150
22 75 40 1 6 52 400
23 75 20 1 6 25 61
0
Dose (mg/kg) of orally administered drug required to
achieve the desired effect.

superior to the first product, the methylamino derivative

13. It was therefore decided to intensify the study of this
compound and prepare it for clinical evaluation and
possible introduction. Its pharmacological and psycho-
tropic properties in animals were thoroughly explored, the
toxicological studies were expanded, and an intensive function and particularly the basic substituent, which was
clinical investigation was started and conducted under the the cornerstone of our initial working hypothesis, proved
to be only unnecessary adornments. The only features
energetic direction of Dr. L. Hines.
It soon became apparent that this compound possessed which were common to these biologically active compounds
were the 1,4-benzodiazepine ring system bearing a chlorine
very valuable tranquilizing (anxiolytic) properties and the
interest of the clinical investigators became so great that at the 7 position and a phenyl group at the 5 position.
within a short time thousands of patients had been treated Based on this knowledge, we started a broad program
with this drug. These extended successful studies enabled of molecular modification, aiming at the discovery of
us to file an NDA very quickly and to introduce the products which would be superior to chlordiazepoxide.
In order to facilitate our work, we first sought simpler
compound [7-chloro-2-(methylamino)-5-phenyl-3H-1,4-
methods which would make compounds of type 21 and 22
benzodiazepine 4-oxide] in 1960 under the trademark
Librium. The generic name which was later generally more readily accessible.
We found that IV-oxides of type 21 could be prepared
accepted was chlordiazepoxide. The time elapsed between
the first pharmacological testing and introduction was only easily by alkaline treatment of 2- (chloromethyl)quinazoline
N- oxides 18.
2V2 years.
This record time was made possible by minimizing red 18
tape and by the enthusiasm and frictionless cooperation
of all the people involved in the chemical and pharma-
ceutical production and pharmacological, toxicological, and
clinical testing of the new drug. The favorable clinical
results and the then existing positive attitude of the FDA
were obviously of prime importance.
While this product was being prepared for introduction,
it became desirable to find a form which would lend itself
to the preparation of a pharmaceutically acceptable elixir
or syrup for pediatric and geriatric use, because chlor-
diazepoxide hydrochloride, the water-soluble clinically used
salt, was extremely bitter. This was not at all surprising
since it is well known to most medicinal chemists that The results of the pharmacological study of these N-
every useful drug is either bitter, hygroscopic, or unstable. oxides (24) were not very interesting, since different
Since this compound was rather valuable, it possessed all substituents had only minimal effect on the biological
three of these properties. spectra and potencies of these compounds. Chemically,
During these studies, we found that a suspension of the however, they undergo an interesting reaction on treatment
quite insoluble, finely pulverized base itself was unsuitable. with acetic acid or anhydride. This so-called Polonovsky
Moreover, the pharmacologically equipotent acetyl de- rearrangement results in the formation of the 3-acetoxy
rivative 20 also proved to be too bitter despite its low derivative 25 which on mild hydrolysis yields the bio-
solubility. Not unexpectedly, aqueous solutions or sus-
AcgO
pensions of 13 and 20 were relatively unstable. The en- 21
suing study led to the very interesting finding that the
substituent at the 2 position was the cause of the instability
and was readily removed by acid hydrolysis.13 To our
pleasant surprise, the decomposition product 21 showed
the same pharmacological activity as 13 (Table II).14 A
further transformation of this product was the removal of
the N-oxide function to form 22. This change also did not
affect the pharmacological properties; quite the contrary,
the activity even seemed to be slightly enhanced. Thus,
it turned out that some of the unique features which
seemed so characteristic for chlordiazepoxide were not at logically active 3-hydroxy derivative 26. This rear-
all needed for its pharmacological activity. The Ar-oxide rangement of benzodiazepine 4-oxides was studied by
Award Address
Table III. Comparison of the Pharmacological Activity0 of Chlordiazepoxide with That of Diazepam
inclined foot
compd screen shock
chlordiazepoxide 100 40
diazepam 30 10
°
Dose (mg/kg) of orally administered drug required to achieve the desired effect.
research teams at Wyeth15 and also at Roche.6
The Wyeth investigators were rather successful. They
discovered the biological activity first, obtained a patent,
and were able to introduce oxazepam in the United States
in 1965 under the trade name Serax and under various
other names in other countries.
We concentrated our studies on simple benzodiazepi-
nones without the N-oxide function. The search for al-
ternative syntheses of these relatively simple compounds
resulted in a number of routes leading to the desired
products in good yields. Two methods which were used
most extensively are shown in Scheme II.16
In both cases, o-amino ketones 27 were used as starting
materials. Treatment of the appropriately substituted
aminobenzophenone with a haloacetyl halide yielded a
compound of type 28, which, on treatment with ammonia,
gave the benzodiazepinone 30 via an amino derivative 29.
The other method involved treatment of an amino-
benzophenone with an amino acid ester hydrochloride in
pyridine leading directly from 27 to 30. The first, mul-
tistep method generally gave better overall yields, up to
70-80% of very pure products. The second method fa-
cilitated the synthesis of benzodiazepinones bearing
substituents at position 3, since many -amino acids
bearing a variety of substituents at the a carbon are
commercially available. With these two methods, we
prepared first a number of benzodiazepinones bearing
numerous substituents at different positions in ring A. A
subsequent modification was the introduction of a sub-
stituent in position 1 which was readily achieved by
treatment with base and an alkylating agent.
Near the end of 1959, just before the introduction of
Librium, we were very much aware of its clinical value and
started to look for a superior product. All the compounds
which were then on hand had similar activity spectra, but
one, the 1-methyl derivative 31 (7-chloro-l,3-dihydro-l-
31 (diazepam, Valium)
methyl-5-phenyl-2/i-l,4-benzodiazepin-2-one), was sig-
nificantly more potent than chlordiazepoxide. In the hope
that this higher potency would be connected with other
advantages in its clinical utility, we started an intensive
study of this substance. The compound was 3-10 times
as potent as chlordiazepoxide, as Table III indicates.
Further studies showed that the toxicity was extremely
low. Its psychotropic and other pharmacological properties
were studied in depth with very favorable results. It was
given the generic name diazepam and, after the appro-
priate toxicological and extended clinical studies, was
introduced near the end of 1963 under the trademark of
Valium. In this case, the time elapsed between the first
Journal of Medicinal Chemistry, 1979, Vol. 22, No. 1 5
anticonvulsant tests
electroshock
pentylene-
cat tetrazole max min
2 18 92 150
0.2 1.4 6.4 64
Scheme II
R
R
NHCOCH-Hal
R
NH2
CO
Hoi -
I
COCH-Hol
R
f CO
pharmacological testing and introduction was 4 years.
The valuable clinical properties of this new compound
led to an expansion of our synthetic program. The
chemical staff was enlarged considerably and the Phar-
macological Department grew proportionally.
As soon as the Librium patent12 had appeared, other
research centers also started the investigation of benzo-
diazepine derivatives. This intensive activity led to a
number of valuable alternative routes for the synthesis of
benzodiazepinones.
Within a few years our intensive efforts resulted in the
synthesis and pharmacological evaluation of well over 3000
1,4-benzo- and heterodiazepinones. This involved the
preparation, identification, and pharmacological investi-
gation of about 4000 intermediates and byproducts.
The large number of benzodiazepinones at our disposal
enabled us to study thoroughly the structure-activity
relationships in this series. It became apparent at the very
beginning of our studies that the substitution pattern
played an important role, of paramount importance being
the substituent at the 7 position (ring A). Substituents
in rings C and also B had additional effects. Our most
significant findings are summarized in Chart I17 (see also
ref 18).
These “rules” proved to be valuable guidelines in the
course of our further studies, which led to benzodiazep-
inones with over 80 different substituents at the 7 position
and with hundreds of substituents at position 1. Of
particular interest was the 1-terí-butyl homologue19 of
diazepam, which is almost completely inactive. Whereas
other alkyl groups are readily removed by liver microsomes,
the ferf-butyl group is not attacked, as was shown by our
Metabolism Group under the direction of Dr. M.
Schwartz.20 Based on these findings we synthesized a
compound which combined all the features known to
6 Journal of Medicinal Chemistry, 1979, Vol. 22, No. 1 Sternbach

Chart I. Effects of Substituents on the Biological ch art II

Activity of Benzodiazepinones0

chlordiazepoxide0 diazepam
(Librium, 1960) (Valium, 1963)
0
Ring A: (position 7) generally, increased by electron-
withdrawing groups, e.g., halogens, N02, and CF3, and
decreased by electron-releasing groups such as CH3 and
OCH3; decreased by any substituents in any positions
other than 7. Ring B: increased by a methyl group at
position 1; decreased by larger substituents; ferf-butyl
derivative is completely inactive. Ring C: increased by
halogens at the 2' position (e.g., Cl and F); very strongly
decreased by a substituent at the 4’ position.
oxazepam
impart high activity: the CH3 group at 1, the nitro group (Serax, 1965)
at 7, and a fluorine at 2'.14,17 It proved to be, as expected,
one of the pharmacologically most potent benzodiazepi-
nones, illustrating the additive or potentiating properties
of pharmacophoric groups in the benzodiazepine series. It COOK·KOH
was introduced in Switzerland in 1975 as a potent hyp-
notic, acting in 1-mg doses.
These studies showed that our six preliminary tests gave
a good indication of the potency of these compounds.
However, differences in the pharmacological spectrum were clorazepate clonazepam
not very significant since, to date, they unfortunately have (Tranxene, 1972) (Clonopin, 1975)
not led to compounds which show effects going much
beyond those of other 1,4-benzodiazepines. In every case,
muscle relaxation and sedation, anxiolytic, anticonvulsant,
and hypnotic properties are present to a varying degree.
Only the preponderance of one or the other activity seems
to vary. It should be noted that this discussion has been
limited to benzodiazepinones having a phenyl group at the
5 position, because such compounds were generally bio-
logically most active and also most readily accessible.
However, many benzodiazepinones bearing other sub-
lorazepam prazepam
stituents at the 5 position were synthesized and studied (Verstran, 1977)
(Ativan, 1977)
pharmacologically by many research teams, including ours.
Few were of interest and some of them were rather difficult 0
Marketed mainly as the hydrochloride. 6 Marketed
to prepare. Only two benzodiazepinones bearing a sub- as the hydrochloride.

stituent other than phenyl at the 5 position are currently

on the market. One is an «-pyridyl derivative, broma-
zepam,21 marketed by Roche; the other is a cyclohexenyl
derivative, tetrazepam,22 which was introduced by Clin-
Byla.
Research in the benzodiazepine series has been very
active and continues in many industrial centers, leading
to various modifications of the basic structure. Some of
the most interesting novel developments are derivatives
with additional rings joining the diazepine nucleus at the
1 and 2 positions. At the present time, the most interesting
are the “triazolobenzodiazepines” 33, which were syn-
thesized and investigated by Takeda23 and Upjohn24 re-
search teams. These compounds are generally more potent
than the corresponding 1-methylbenzodiazepinones. Two
of them are commercially available outside the U.S.
Evidence of the intensive research in the benzodiazepine
field still in progress is indicated by the continuous flow
of new patents and scientific publications. In the last five
years, over 1600 original patents have appeared, and over
12000 papers concerned with the chemistry, pharmacology,
and clinical aspects have been published. One of the main
objectives is the discovery of products having a narrower
spectrum of biological activities. It is to be expected that
this search will continue for years to come and might
ultimately lead to superior products with more specific
properties: anxiolytics, muscle relaxants, or anticonvul-
sants causing less sedation, compounds with pronounced
antidepressant properties, or even drugs acting in psy-
choses. The general acceptance of this class of compounds
is illustrated by the widespread use of the eight benzo-
diazepine derivatives which are now marketed in the U.S.
They are shown in Chart II together with their generic
trade names and introduction dates. Six of them are
[1 -Penicillamine,2- leucine] oxytocin Journal of Medicinal Chemistry, 1979, Vol. 22, No. 1 7

anxiolytics, flurazepam is a hypnotic, and clonazepam is
an antiepileptic. Fourteen additional 1,4-benzodiazepine
derivatives are marketed outside the U.S., many of them
under several trade names.
Acknowledgment. I wish to express my appreciation
to all my associates who contributed so greatly to the
development of the chemistry of benzodiazepines, in
particular to the first group of chemists who were involved
in the earliest stages of our studies: G. A. Archer, . E.
Derieg, G. F. Field, R. I. Fryer, W. Metlesics, R. Y. Ning,
E. Reeder, G. Saucy, R. A. Schmidt, N. Steiger, and A.
1961); L. O. Randall, W. Schallek, G. A. Heise, E. F. Keith,
and R. E. Bagdon, J. Pharmacol. Exp. Ther., 129,163 (1960).
(8) For a description of these tests see ref 17.
(9) L. H. Sternbach and E. Reeder, J. Org. Chem., 26, 1111
(1961).
(10) L. H. Sternbach, Angew. Chem., 83, 70 (1971).
(11) L. H. Sternbach, E. Reeder, O. Keller, and W. Metlesics,
J. Org. Chem., 26, 4488 (1961), and additional unpublished
results.
(12) L. H. Sternbach, U.S. Patent 2893992 (July 7, 1959).
(13) L. H. Sternbach and E. Reeder, J. Org. Chem., 26, 4936
(1961)
.

(14) L. H. Sternbach and L. O. Randall, CNS Drugs, Symp.,

Stempel. Thanks are also due to our co-workers in Basle,
Switzerland, under the direction of Dr. J. Hellerbach. I
also wish to thank Dr. L. 0. Randall and his staff for their
splendid cooperation.
Hyderabad, India, 53 (1966).
(15) S. C. Bell and S. J. Childress, J. Org. Chem., 27,1691 (1962).
(16) L. H. Sternbach, R. I. Fryer, W. Metlesics, E. Reeder, G.
Sach, G. Saucy, and A. Stempel, J. Org. Chem., 27, 3788
(1962) .

References and Notes (17) L. H. Sternbach, L. O. Randall, R. Banziger, and H. Lehr,

(1) This is a condensed version of the Medicinal Chemistry
Award Address presented at the 16th Medicinal Chemistry
Symposium in Kalamazoo, Mich., June 20, 1978.
(2) K. Dziewoñski and L. Sternbach, Bull. Int. Acad. Pol. Sci.
Lett., Cl. Sci. Math. Nat., Ser. A, 416 (1933) [Chem. Abstr.,
28, 2717 (1934)]; ibid., 33 (1935) [Chem. Abstr., 30, 2971
(1936)].
(3) K. Auwers and F. von Meyenburg, Chem. Ber., 24, 2370
(1891).
(4) J. Meisenheimer and A. Diedrich, Chem. Ber., 57, 1715
(1924); K. von Auwers, ibid., 57, 1723 (1924).
(5) L. H. Sternbach, S. Kaiser, and E. Reeder, J. Am. Chem.
Soc., 82, 475 (1960).
(6) E. Reeder and L. H. Sternbach, unpublished results.
(7) L. O. Randall, Dis. Nerv. Syst., Suppl. 7, 22, Sect. 2 (July
“Medicinal Research Series”, Vol. 2, A. Burger, Ed., Marcel
Dekker, New York, N.Y., 1968, p 237. See also ref 14.
(18) S. J. Childress and . I. Gluckman, J. Pharm. Sci., 55, 577
(1964).
(19) N. W. Gilman and L. H. Sternbach, J. Heterocycl. Chem.,
8, 297 (1971).
(20) M. Schwartz, unpublished results.
(21) R. I. Fryer, R. A. Schmidt, and L. H. Sternbach, J. Pharm.
Sci., 53, 264 (1964).
(22) J. Schmidt, P. Comoy, M. Suquet, J. Boitard, J. LeMeur,
J. J. Basselier, M. Brunaud, and J. Salle, Chim. Ther., 2,
254 (1967).
(23) K. Meguro and Y. Kuwada, Tetrahedron Lett., 4039 (1970).
(24) J. B. Hester, Jr., D. J. Duchamp, and C. G. Chidester,
Tetrahedron Lett., 1609 (1971).

[l-Penicillamine,2-leucine]oxytocin. Synthesis and Pharmacological and

Conformational Studies of a Potent Peptide Hormone Inhibitor1
Victor J. Hruby,* K. K. Deb, Diane M. Yamamoto,2
Department of Chemistry, University of Arizona, Tucson, Arizona 85721
Mac E. Hadley,
Department of General Biology, University of Arizona, Tucson, Arizona 85721
and W. Y. Chan
Department of Pharmacology, Cornell University Medical College, New York, New York 10021. Received May 10, 1978

[1-Penicillamine,2-leucine] oxytocin was synthesized by the solid-phase method of peptide synthesis and purified
by partition chromatography on Sephadex G-25, followed by gel filtration. The peptide was found to be a very
potent competitive inhibitor of oxytocin in the oxytocic assay with a pA2 of 7.14 and an inhibitor of oxytocin in
the milk-ejecting assay. The compound showed no agonist activity in either of these assays, and its inhibitory activity
at the uterus was of prolonged duration. The 13C nuclear magnetic resonance spectral properties and the 13C \
(spin-lattice) relaxation times of [Pen1,Leu2]oxytocin were determined, and the results were compared with previous
studies of [Pen^oxytocin, a related competitive inhibitor, and oxytocin, the native hormone agonist. These studies
indicated that the hormone inhibitors [Pen1,Leu2] oxytocin and [Pen1] oxytocin have similar conformational and dynamic
properties which are different than those of the agonist, oxytocin.

Peptide hormone competitive inhibitors (antagonists)
constitute a potentially useful class of organic compounds
in clinical applications and for studying peptide-receptor
interactions and the mechanisms of peptide hormone
action. These applications derive from their ability to
interact with the receptor in a manner similar to the
hormone and their inability to transduce a biological
message to effect a change in the target cells metabolism
or other properties. Thus a peptide hormone competitive
inhibitor can provide information of the hormone-receptor
interaction independent of the transduction event and
important clues to structural and dynamic features related
to both binding and transduction.
Recently we have shown that [l-penicillamine]oxytocin
([Pen1] oxytocin, S-C(CH3)2CH(NH2)CO-Tyr-Ile-Gln-
Asn-Cys-Pro-Leu-Gly-NH2), a competitive inhibitor of
oxytocin,3·4 has considerably restricted dynamic properties
relative to those of oxytocin.5’6 These studies suggested
that certain specific differences in the conformational,
dynamic, and structural properties of the hormone and
antagonist were related to differences in biological activity.6

0022-2623/79/1822-0007$01.00/0 © 1978 American Chemical Society